Open AccessMACHINE-LEARNING ALGORITHM FOR SHIELDED SPECIAL NUCLEAR MATERIALS DETECTION
Author(s) -
Eugene Masala,
L. Blomeley
Publication year - 2019
Publication title -
cnl nuclear review
Language(s) - English
Resource type - Journals
eISSN - 2369-6931
pISSN - 2369-6923
DOI - 10.12943/cnr.2018.00004
Subject(s) - overfitting , computer science , algorithm , convolutional neural network , artificial neural network , artificial intelligence , pooling , machine learning , detector , telecommunications
A machine-learning algorithm has been implemented by use of a neural network as a preliminary study on the applicability of this method to special nuclear materials detection. The algorithm predicts the presence of the 238 U isotope when learning from a gamma spectrum data measured with a high-purity germanium detector from a sample of depleted uranium. In this work, both a fully connected neural network and a convolutional neural network have been implemented, and the performance of different configurations of the network has been studied. The use of convolutional network showed better performance over the fully connected network, with cost function and success rate values supporting a better prediction while avoiding overfitting. Furthermore, implemented network features such as filtering, max-pooling, dropout regularization, and momentum optimization also showed improved prediction performance.